BR112016011660B1 - SEPARATION MATRIX, AND, METHODS FOR STABILIZING A FERMENTED BEVERAGE AND FOR MANUFACTURING THE SEPARATION MATRIX - Google Patents

Abstract

separation matrix, and methods for stabilizing a fermented beverage and for manufacturing the separation matrix. the invention describes a separation matrix with a porous solid support and a plurality of polyvinylpyrrolidone (pvp) polymer chains covalently attached to the solid support. polyvinylpyrrolidone polymer chains are either vinylpyrrolidone copolymers or homopolymer chains comprising at least 70% by mole of vinylpyrrolidone monomer residues and less than 2% by mole of negatively charged monomer residues.

Description

Technical field of invention

[001] The present invention relates to the stabilization of beverages, and more particularly to separation matrices for the stabilization of fermented beverages. The invention also relates to a method for stabilizing beverages and a method of manufacturing a separation matrix for stabilizing fermented beverages.

Fundamentals of the invention

[002] In fermented beverages such as beer and wine, undesirable haze can form during storage due to precipitation/aggregation of active compounds in haze (HA). These compounds may comprise HA polyphenols and HA polypeptides, as well as complex reaction products between them. To increase the colloidal stability of beverages and enhance their shelf life, the beverage can be treated to reduce concentrations of HA compounds, which is said to stabilize beer or wine.

[003] There are several technologies to clarify beer and wine in order to improve the colloidal stability of turbidity. The most common means of removing haze precursors is to add silica hydrogel (SHG) and/or cross-linked polyvinylpyrrolidone particles (polyvinylpolypyrrolidone or PVRP) into unstabilized beer. Dosages are typically 1-40 g PVRP/hl and ~50 g SHG/hl (CW Bamforth: J Am Soc Brew Chem 57(3), 81-90, 1999). SHG is designed to reduce turbidity-active polypeptides and PVRP to reduce polyphenols. Both SHG and PVPP are used in the form of fine particles added to the beverage and then removed by filtration. Traditionally filter cakes have been discarded, but methods for regenerating PVPP particles in a separate regeneration facility have also been suggested, as described in, for example, WO 2012/011808. This will, however, require complex operations and space for additional equipment. Pat. US 6,001,406 describes a method for stabilizing beverage using an ion exchanger, in particular a water-insoluble porous hydrophilic matrix to which ion exchange groups are covalently bonded. Such a system has been marketed by Handtmann Armaturenfabrik GmbH & Co. KG and GE Healthcare BioSciences AB under the name Combined Stabilization System (CSS), using crosslinked positively charged agarose beads in a packed bed column. The beads remain on the spine during both stabilization and regeneration cycles and can be reused for years without any extra manipulation.

[004] With increased demands regarding process efficiency and beverage stability, there is, however, a need for adsorbents with improved binding capacity for HA compounds. Pat. US 8,137,559 describes an attempt to achieve this with diethylene glycol based binders on agarose beads, but there is still a need for adsorbents that show improved stabilization and that are able to stabilize higher volumes of beverage prior to regeneration.

Summary of the invention

[005] One aspect of the invention is to provide a separation matrix with improved binding capacity for HA compounds in fermented beverages. This is achieved with a separation matrix as defined in claim 1.

[006] One advantage is that the separation matrix is able to improve the haze stability of fermented beverages compared to prior art matrices. Additional advantages are that the separation matrix can be used and regenerated in packed beds, because it does not produce any leachables that are of concern for food purity or food safety, and that the matrix can be easily manufactured.

[007] A second aspect of the invention is to provide a method of stabilizing fermented beverages, allowing for an increased loading of beverage in the matrix. This is achieved with a method as defined in the claims.

[008] A third aspect of the invention is to provide a process for making a separation matrix with improved capacity for HA compounds in fermented beverages. This is achieved with a method as defined in the claims.

[009] Additional appropriate embodiments of the invention are described in the dependent claims.

Definitions

[0010] "Vinyl pyrrolidone" here means l-vinyl-2-pyrrolidinone, CAS 88-12-0, which is also commonly known as N-vinyl pyrrolidone. figures

[0011] Fig. 1 shows examples of unique linker moieties for PVP chains: a) derived from allylic groups and b) derived from thiol groups. S is support.

[0012] Fig. 2 shows a grafting process of the invention, with a hydroxy-functional support first reacted with allyl glycidyl ether (AGE) and vinyl pyrrolidone (VP) then grafted onto the allyl-functional support.

[0013] Fig. 3 shows a typical stabilization curve for the anion exchange matrix of the CSS adsorber.

[0014] Fig. 4 shows the application fit of the miniaturized beer stabilization test.

[0015] Fig. 5 shows typical production data from the EBC cold turbidity method.

[0016] Fig. 6 shows the tanoid content in unstabilized beer, reference stabilized beer and beer stabilized with three prototypes of the invention.

[0017] Fig. 7 shows graphical views of the normalized results for the cold turbidity analysis. The three graphs refer to three batches of new beer, produced on three different days.

[0018] Fig. 8 shows the reduction of total polyphenol for the prototypes.

[0019] Fig. 9 shows the correlation between total polyphenol reduction and cold turbidity.

[0020] Fig. 10 shows reaction temperature curves for selected experiments. 311 - 55°C 1.9% initiator; 354 - 55°C 1.0% initiator; 374 - 45°C 1.0% initiator; 392 - 35°C 1.0% initiator; 493 - 45°C 1.0% initiator, 1M Na2SO4.

Detailed description of modalities

[0021] In one aspect the present invention describes a separation matrix comprising a porous solid support and a plurality of polyvinylpyrrolidone (PVP) polymer chains covalently bonded to said solid support. Polyvinylpyrrolidone polymer chains are vinylpyrrolidone copolymer chains or homopolymer chains comprising at least 70% by mole of vinylpyrrolidone monomer residues and less than 2% by mole, or less than 1% by mole of monomer residues. negatively charged. Negative charges are likely to repel the mostly negatively charged HA compounds and thus reduce matrix performance. It may also be advantageous if the polyvinylpyrrolidone polymer chains are free of negatively charged monomer residues. Suitably, the polyvinylpyrrolidone polymer chains are not cross-linked in order to provide high accessibility and mobility to the HA compounds.

[0022] In some embodiments, the polyvinylpyrrolidone polymer chains comprise up to 30% by mole of positively charged monomer residues. Positive charges can give additional advantageous interactions with negatively charged HA compounds. Positively charged monomer residues can be introduced by vinyl pyrrolidone graft polymerization with comonomers such as, for example, diallyldimethyl ammonium chloride (DADMAC), 3-(methacryloylamino)propyltrimethyl ammonium chloride (MAPTAC), 2-(methacryloyloxy) chloride )ethyltrimethylammonium, vinylpyridine, etc.

[0023] Polyvinylpyrrolidone polymer chains may alternatively or additionally comprise uncharged monomer residues other than vinyl pyrrolidones. Such residues can, particularly if they contain hydrophobic and/or hydrogen bonding moieties, favorably modulate interactions between the chains and the HA compounds. Specific monomers that can be used for this purpose include, for example, N-vinyl caprolactam, styrene, alkyl vinyl ethers, (hydrophobic) and, for example, PEG-monomethacrylates, PEG-monovinyl ethers, PEG-monoalkyl ethers, hydroxyalkylvinyl ethers, allyl alcohol (hydrogen bond). If only uncharged comonomers are used, their amount can be up to 30% by mole, and if a combination of positively charged and uncharged comonomers is used, their combined amount can be up to 30% by mole.

[0024] In some embodiments the solid support may comprise positively charged moieties that are not part of the polyvinylpyrrolidone polymer chains. Such moieties can, for example, be introduced by reacting hydroxyl groups on the support (for example, a polysaccharide support such as agar or agarose) with positively charged reagents. Examples of such reagents include glycidyltrimethylammonium chloride and diethylaminoethyl chloride.

[0025] In certain embodiments, the porous solid support comprises particles, such as particles having an average diameter (heavy volume) of 10-500 microns. Large particles give lower back pressures in the packed beds and for this purpose it may be advantageous if the particles have an average diameter of 150-500 microns, such as 200-400 microns. The particles may suitably be spherical or substantially spherical, which facilitates filling.

[0026] In some embodiments, the porous solid support has a porosity of 80-98%, such as 90-98%. High porosity is advantageous because the support can accommodate a large amount of polyvinylpyrrolidone polymer chains without blocking the pores. Porosity is defined as the volume fraction of pores in the support and can be conveniently measured by measuring the solids content of a support balanced with distilled water. With membrane-formed or monolithic supports, the water-balanced support is free of excess water, weighed and dried in an oven at eg 100oC and reweighed. The pore volume fraction can then be calculated using an estimate of the density of the pore wall material. Particle-shaped supports are freed of excess water by gentle vacuum suction on a glass filter until a particulate filter cake is formed. Excess water is removed when the water level has receded below the top of the filter cake until a first crack in the filter cake is formed. Prolonged suction should be avoided to avoid evaporation in the filter cake. A sample of the filter cake is then weighed and dried and the porosity is calculated from the wet (mwet) and dry (mdry) weights, using an estimate of the density of the pore wall material. Typical densities of pore wall material (pmat) are: agarose 1.5 g/cm3, styrene-divinylbenzene 1.1 g/cm3, methacrylate polymers 1.15 g/cm3, polyvinyl alcohol 1.2 g/cm3, silica 2.2 g/cm 3 . The porosity (p) in % by volume can then be calculated as: p = 100 * ((mwet - mdry/ph20)/(( mwet - mdry)/ph20 + mdry / pmat) where ph20 is the density of water in the temperature in question, typically 1.0 g/cm3.

[0027] Alternatively, the pore structure of the support can be characterized by inverse size exclusion chromatography, in which case a Kav value is obtained. The Kav value is a measure of the volume fraction of the support that is accessible for a probe molecule of a given size.

[0028] Typically protein molecules with well defined sizes are used as probe molecules, but it is also possible to use dextran fractions as probe molecules. Details of measurements and calculations of Kav values are given in Gel Filtration Principles and Methods, Pharmacia LKB Biotechnology 1991, particularly p. 10-11, which is incorporated herein by reference. In some embodiments, Kav for human serum albumin (Mw 67 kDa) on the support is at least 0.4, such as at least 0.5 or 0.5-0.9.

[0029] In certain embodiments, the porous solid support comprises a polymer selected from the group consisting of styrenic polymers, methacrylate polymers, vinyl ether polymers, vinyl alcohol polymers and polysaccharides.

[0030] In some embodiments, the porous solid support comprises a polysaccharide selected from the group consisting of agarose, agar, cellulose and dextran. Polysaccharides are hydrophilic, which minimizes the risk of fouling due to hydrophobic interactions with compounds in the beverage. Agarose and agar can be easily prepared as high porosity (eg 90-98% porosity) hydrogels with high rigidity by thermal gelling. Of the two, agarose is more expensive but has the advantage that it is essentially free of negative charges. If agar is used, it can be advantageously treated with alkali to remove hydrolyzable negatively charged groups.

[0031] Polysaccharide supports may suitably contain less than 10 or less than 5 micromol/ml negatively charged and acidic groups. This content can be determined by titration methods well known in the cation exchanger art.

[0032] The supports may additionally comprise extenders, that is, polymers connected to the surface of the pores of the support. The extenders can be, for example, hydroxy-functional polymers, in particular polysaccharides of the dextran type, which can be connected by covalent coupling on the hydroxyl groups of the supports. A specific example is agar or agarose beads with dextran polymers attached over agar/agarose hydroxyls that have been activated by epoxy using epichlorohydrin or a diepoxide. The extenders may be appropriately connected by linking structures comprising ether groups which are stable under alkaline regeneration of the matrices. Extenders can facilitate interactions between HA compounds and PVP chains and can increase mass transport rates.

[0033] In certain embodiments a plurality of polyvinylpyrrolidone polymer chains are each covalently linked to said solid support by a single binding moiety. The single bonding moiety may comprise, for example, an ether-linked C3 chain, such as in, -O-CH2CHCH2- or -O-CH2-CH(OH)-CH2-O-CH2CHCH2- as illustrated in Fig. 1 a). Such linking moieties can be achieved by reacting hydroxyl groups on a support with an allyl halide or allyl glycidyl ether and polymerizing vinyl pyrrolidone in the presence of the support. Two ends of the polyvinylpyrrolidone polymer chain can be joined by the same linking moiety if the polymerization proceeds through an allyl group. In this case, the matrix may additionally comprise a few polyvinylpyrrolidone polymer chains which are linked by two linking moieties, if the allyl density is high enough such that one chain may polymerize through two adjacent allyl groups. The single bonding moiety may also comprise a thioether bond, for example as illustrated in Fig. 1b). In this case, thiol groups were introduced on the support and these have acted as chain transfer agents in the polymerization of vinylpyrrolidone. Binding by a single binding moiety, as opposed to multi-point attachment, improves the mobility of the polymer chains and thus their accessibility and interaction with HA compounds.

[0034] In some embodiments, the matrix comprises 0.50-4.0, such as 0.70-3.0 or 1.0-3.0 micromoles of vinylpyrrolidone monomer residues per ml of matrix. The content of vinylpyrrolidone monomer residues can be determined spectroscopically (eg NMR OR FTIR) or, for example, by nitrogen elemental analysis, where any nitrogens introduced by positively charged monomers can be titrated and subtracted from the total nitrogen. A high content of vinylpyrrolidone monomer residues is advantageous for both binding ability and stability of the treated beverage.

[0035] In certain embodiments, the matrix comprises 0.50 - 0.80 g, such as 0.60-0.80 g of polyvinylpyrrolidone polymer per g of dry matrix. Alternatively, the matrix may comprise 100-200 mg, such as 120-180 mg of polyvinylpyrrolidone polymer per ml of drained matrix.

[0036] A high content of polyvinylpyrrolidone polymer is advantageous for both binding capacity and stability of the treated beverage. High contents of polyvinylpyrrolidone polymer are particularly advantageous in combination with high porosities as discussed above. Excessively high contents of polyvinylpyrrolidone polymer can, however, have a negative influence on the mass transport and/or the mechanical properties of the matrix.

[0037] In some embodiments, the matrix comprises less than 5 micrograms, such as less than 2 or less than 1 microgram of leachable carbon per g of dry matrix. Low leachables can be achieved by carefully washing the matrix before use and by using support materials which themselves have low leachables. The amount of leachable carbon can be properly determined by the methods described in M Andersson ET AL: Process Biochemistry 33(1), 47-55, 1998, in which 10 ml of swollen matrix water is incubated in 50 ml of high purity water. for 1 week and the total organic carbon (TOC) content of the water supernatant is determined. It is advantageous to minimize the content of leachables/extractables due to food safety concerns, but also in relation to purity regulation concerns, eg beer in certain countries.

[0038] Poly(vinyl pyrrolidone) (in the following parts denoted PVP) can be fixed in a graft reaction sequence as shown in Figure 1.

1. Allylation.

[0039] In this step allila groups are introduced. Allyl glycidyl ether or an allyl halide is fixed under alkaline conditions in a hydroxyl-functional base matrix (for example, a cross-linked agarose base matrix such as Sepharose™ Big Beads).

2. PVP graft.

[0040] Vinyl pyrrolidone (VP) and a radical initiator (eg ADBA) are dissolved in an aqueous solution containing the allylated particles. Pure water or aqueous salt solutions (eg sodium sulfate) are examples of possible reaction solvents. The mixture is heated, whereby the initiator decomposes and forms free radicals that initiate a polymerization of vinyl pyrrolidone. Developing PVP chains can also react with the allyl groups on the particles and thereby PVP chains are covalently attached to the particles.

[0041] The fine structure of the PVP chains is not known in detail, for example the molecular weight, molecular weight distribution and how many of the allyl groups are consumed.

[0042] A main parameter that can be measured is the dry weight (solids content) of 1 mL of gel before and after the grafting reaction. In this way the amount of PVP fixed in mg/ml can be calculated.

[0043] Other parameters such as average amount of PVP/allyl group (which gives an indication of average chain length) or % PVP in the structure can also be calculated.

[0044] The following parameters are believed to affect the amount of PVP that is set. • The concentration of VP in the reaction mixture. • Initiator concentration • Polymerization temperature • Amount of allyl groups attached to particles • Type of initiator • Reaction solvent • Slurry concentration of particles (the amount of allylated particles) • Base matrix parameters (particle size, distribution pore size, dry weight)

[0045] The polymerization of VP is an exothermic reaction. The potential temperature rise will be affected by the rate of the reaction, the total amount of heat, and how much reaction solvent is present to absorb the heat (ie, the heat capacity of the solvent).

[0046] The total amount of heat generated will be governed primarily by the amount of VP used, while the rate of reaction will determine the amount of heat generated per unit of time. Initiator concentration and reaction temperature are the main parameters that will affect the reaction rate. A temperature log can be used to track the temperature in the reaction vessel over time.

[0047] In one aspect the present invention describes a method for stabilizing a fermented beverage, for example beer, comprising the steps of: a) providing a column packed with a separation matrix comprising a porous solid support and a plurality of chains of polyvinylpyrrolidone polymer covalently attached to said solid support, wherein said polyvinylpyrrolidone polymer chains are copolymer chains or vinylpyrrolidone homopolymer chains comprising at least 70% by mole of vinylpyrrolidone monomer residues. The separation matrix may, for example, be a separation matrix as described above; b) passing the beverage through said column and recovering a flow through the column as a stabilized beverage. In step b), the residence time of the beverage in the column can be, for example, 2 minutes or less, such as 1 minute or less or 10 s - 1 min. The residence time is calculated as the bed height (cm) of the column divided by the flow rate (cm/min) of the beverage through the bed of the column. A short residence time is desirable for economic reasons and the matrices of the invention have a sufficiently high mass transport rate to allow sufficient stabilization and turbidity with residence times below 1 or 2 minutes. The stabilized beverage (beer) in step b) may have a cold haze of less than 25% of the cold haze for the beverage before passing through said column. Cold turbidity can be properly measured according to the well-known EBC method (method no. 31.1) and the residence time in the column can be less than 1 minute, such as 18 s.

[0048] In certain embodiments the method additionally comprises the steps of: c) regenerating the column with a regeneration solution and; d) repeating steps a) - c) at least twice, such as at least 10, at least 50 or at least 500 times. This has the advantage that the separation matrix is reused in several cycles, which is beneficial to the overall economy of the process.

[0049] In some embodiments the regeneration solution comprises NaOH, such as at least 0.1M NaOH or 0.1-2M NaOH, for example about 1M NaOH. NaOH solutions can efficiently remove adsorbed contaminants and NaOH is also an accepted beverage industry cleaning agent, which if the pH is controlled after cleaning does not leave any unpleasant tasting or toxic residues.

[0050] In a third aspect the present invention describes a method of manufacturing the separation die as described above. This method comprises the steps of: a) providing a porous solid support comprising at least 5 micromol/ml of radical reactive moieties. The radical reactive moieties may be polymerizable moieties such as C=C double bonds, chain transfer moieties such as thiols or they may be immobilized initiators; b) contacting the support with a monomer composition wherein at least 70 mol % of the monomers in said monomer composition is N-vinyl pyrrolidone and less than 1 mol %, or less than 2 mol % of the monomers are negatively charged; c) initiating free radical polymerization to form a matrix having polyvinylpyrrolidone polymer chains covalently attached to the support, and; d) washing the matrix.

[0051] In certain embodiments step a) comprises either i) providing a divinyl benzene copolymer support comprising at least 5 micromol/ml residual double bonds, or ii) reacting a hydroxyl-functional support with an allyl halide or allyl glycidyl ether.

[0052] In some embodiments step c) is carried out with the support suspended in an aqueous solution comprising the monomer composition and 0.05-3 mol/l of a salt, such as 0.1 - 2 M of sodium sulfate. sodium or ammonium sulfate. An advantage of using an aqueous salt solution is that it has a high heating capacity and that the amount of monomer needed to have a given amount of PVP grafted can be reduced. These factors mean that temperature rise due to exothermic heat generation can be avoided, which is important for process scalability.

Alcohol cold cloudiness

[0053] Description of the Pfeuffer GMBH assay. European brewers convention (EBC) analytical method no 31.1.

[0054] "When very chilled beers demonstrate a reversible turbidity which is dependent on the condition of the beer and is caused by precipitated polyphenol protein complexes. The addition of ethanol reduces the solubility of the complexes and thereby accelerates the formation of turbidity. The low temperature test, which can be performed quickly, makes it possible to predict the long-term turbidity of the beer. Even immediately after the beer stabilization treatment, this test provides accurate information on the beer's turbidity potential and the effectiveness of the stabilization measures. , which can then be evaluated and changed if necessary.

Description of the Pfeuffer GMBH essay.

[0055] "Tanoids are the amount of polyphenols that are PVP precipitable. Among them are low to medium molecular weight polyphenols, as well as catechin polymers and anthocyanogens. Tanoids come from malt and hops. Although they are present in small amounts in beer, they are of great significance with respect to colloid stability and aroma consistency. The tanoid content of beer, wort extracts, barley, malt, and hops can be determined by precipitation using PVP. PVP, a compound protein-like, it attaches to the tanoids by bridge bonds to H and forms insoluble complexes with them, thereby leading to turbidity. If a PVP solution is continuously added to the sample, the turbidity will increase until all the tanoid molecules are fixed to PVP. An additional batch of PVP will lead to an increase in turbidity due to dilution. The amount of PVP added until peak turbidity has been reached is proportional to the t content anoid".

total polyphenols

[0056] Description of the EBC method test 9.11. The total polyphenol assay covers all polyphenols. Beers with high total polyphenols >200 mg/L are considered to be difficult to stabilize while beers with total polyphenols < 150 mg/L are easier to stabilize using PVPP treatments. The determination of total polyphenol is made by reacting beer polyphenols with iron ammonium citrate under basic conditions. The ferric-polyphenol complex is measured at an absorption at 600 nm against a blank solution.

Examples Examples 1. Study of the graft

[0057] The main purpose of the graft study was to find reaction conditions where an appropriate amount of PVP could be fixed, where the temperature is under control and the boiling point can be avoided. The following synthesis parameters were included in the study. • Polymerization temperature (ranged between 35 and 55°C) • Initiator concentration (ranged between 1.0 and 1.9% w/w) • Amount of allyl groups (ranged between 82 and 170 μmol/mL) • Concentration of VP (ranged between 10.0 and 39.2% weight/weight) • Slurry concentration (ranged between 18.5 and 36.9% volume/weight)

Quantidade de gel [mL] A quantidade de gel a ser enxertada. Concentração da pasta fluida da reação [volume/peso]:[0058] The identity of the initiator (ADBA) can be kept constant from start to finish as long as it is rapidly soluble in water and decomposes in an appropriate temperature range. Table 1. Prototype overview

Amount of Gel [mL] The amount of gel to be grafted. Reaction slurry concentration [volume/weight]:

[0059] This parameter decides how concentrated the reaction slurry should be. A higher value means that there is less reaction solution per mL of gel. % VP + primer [% weight/weight]

[0060] This parameter determines how much of the reaction suspension in % by weight that should consist of VP and initiator, of the active substances. Initiator concentration [% weight/weight]

[0061] This parameter determines how many% of initiator should be used in relation to the amount of VP.

[0062] See typical formulas below.

Allylation method (typical formula)

[0063] 6% agarose beads crosslinked with epichlorohydrin and having an average diameter (volume-heavy) of 200 microns (a sieve fraction between 100 and 300 microns) were used as the base matrix.

[0064] 100 mL of base matrix gel was transferred to a glass filter (pore size ratio 2) and sucked dry. Thereafter the gel was washed with 2000 ml of water in portions. The gel was sucked dry after the last wash and transferred to a 500 ml round bottom flask and water was added to a total weight of 87.14 g. 92.16 g of 50% (wt/wt) sodium hydroxide solution was added.

[0065] The round bottom flask was assembled and immersed in a bath that was maintained at 50°C. A stirrer with two 5 cm diameter oscillating blades was used and the agitation was adjusted to 350 rpm. When the temperature reached 50°C, 33.0 g of allyl glycidyl ether was added. The reaction was allowed to stand for 18 hours. The reaction was then quenched with 114.98 g of 60% acetic acid in portions.

[0066] The gel was washed with 20xGV of water followed by 5xGV of ethanol, 20xGV of water again and finally 3xGV of 20% ethanol solution.

[0067] The allyl content was determined by titration and was found to be 231 micromol/mL.

Grafting method (typical formula), see Table 1 for specific details regarding other syntheses.

[0068] 1.0 M sodium sulfate was prepared by dissolving 284.1 g of anhydrous sodium sulfate in a final volume of 2000 mL in a measuring flask.

[0069] 100 mL of gel = 234.0 g of slurry was transferred to a filter (por. 2) and sucked dry. The gel was washed with 5000 ml of water in portions. Then the gel was washed with 3 X 150 mL of 1.0 M sodium sulfate and sucked dry between washes.

[0070] The round bottom flask was placed on a balance and it was tared. The "dry" gel was transferred to the round bottom flask and the total weight was adjusted to 244.7g with 1.0M sodium sulfate solution.

[0071] 26.3 g of vinyl pyrrolidone and 0.25 g of ADBA were added.

[0072] The round bottom flask has been assembled. A top stirrer motor with a stirrer with two 5 cm oscillating blades was used. Nitrogen gas was bubbled through the reaction solution (with a Pasteur pipette) for about 25 minutes. The stirring rate was 200 rpm.

[0073] A glycerol bath was set to 45 degrees. When this temperature was reached the round bottom flask was immersed in the glycerol bath and the reaction was started.

[0074] After a few hours, a suspension of polymer in water was formed; the polymer particles were about 2 mm in diameter. The reaction was allowed to proceed overnight. ~200 g of distilled water was added to dilute the reaction solution. The polymer lumps disappeared and the reaction solution was easily filtered.

[0075] The gel was washed with 20 L of distilled water, and 1 L of 20% ethanol.

[0076] The dry weight was recorded after the grafting reaction, and was recorded as being 265mg/mL.

dry weight measurement

[0077] A dry weight balance set at 120°C was used for all dry weight measurements. For dry weight measurement 1 mL of gel is transferred from a glass filter with a PTFE top designed to accommodate 1.0 mL of filter cake into the aluminum beaker of the balance.

[0078] A formula is obtained which gives 100-200 mg/mL of fixed PVP, in which there is no significant temperature rise during the reaction and which is in all respects suitable for production scale. Table 2. Overview of additional prototypes.

[0079] The first prototype, 311 had an allyl content of 170 μmol/mL. The amount of PVP that was fixed was 216 mg/ml. The synthesis established for this set of experiments was therefore considered to be OK, although the exotherm caused boiling. A very viscous solution is formed which is spun around the stirrer. The viscosity increased simultaneously with the occurrence of boiling. The reaction, however, never boiled over and was therefore allowed to proceed overnight.

[0080] The next experiment was 354, where the amount of initiator was decreased from 1.9% to 1.0%. The theory was that a lower concentration of initiator would slow down the reaction and this could potentially reduce the risk of boiling. This was not the case; boiling still took place. However, the boiling was actually delayed from 17 minutes to 29 minutes. The amount of PVP that was fixed was actually higher than that of 311, 227 mg/mL compared to 216 mg/mL. For all subsequent experiments the initiator concentration was therefore adjusted to 1.0% as this had a positive effect on both boiling and the amount of PVP that was set. This parameter could, however, be further optimized in future studies, an even lower concentration of initiator could, for example, be interesting to investigate.

[0081] For the next two experiments, 374 and 392, all parameters were kept constant except temperature. Previous experiments were done at 55oC while 45 and 35 degrees were used for 374 and 392 respectively. Going to 45oC, boiling was further delayed until after 59 minutes of reaction. When 35°C was used no boiling occurred. Temperature profiles for these reactions can be found in Appendix A along with selected temperature curves from recent experiments. At 45°C there was even more fixed PVP than at 55°C, 238 mg/mL compared to 227 mg/mL. It may be that the more controlled reaction at 45°C favors PVP fixation. At 35°C, however there is a clear drop in the amount of PVP that is fixed, only 134 mg/mL was fixed in this case. The temperature of the reaction thus had a pronounced effect both on the tendency to boil and on the amount of PVP that is fixed. 45°C was used for all subsequent experiments as at this temperature high amounts of PVP could be clearly fixed, while the tendency to boil was lower than at 55°C.

[0082] In order to totally avoid boiling, the amount of VP in the reaction was divided to 19.6% for the next two experiments. The effect of allyl content was also investigated. For 410 the allyl level was 170 μmol as for all previous experiments, while for 431 the allyl content was 82 μmol/mL. By reducing the VP concentration, boiling was avoided, but the reaction solution still became too viscous and was spun around the stirrer. This issue also had to be resolved in order to have a production friendly formula.

[0083] The amount of PVP fixed was also clearly lower for 410 compared to 374, 120 mg/mL compared to 238 mg/mL. It is interesting to see that a halving of the VP concentration results in half the amount of VP fixed. The effect of the allyl content was not very pronounced. For 431 where an allyl level of 82 μmol/mL was used, 113 mg/mL PVP was fixed compared to 120 mg/mL for 410 where the allyl level was more than twice as high, 170 μmol/mL.

[0084] In order to see if more PVP could be fixed with 196% VP in the reaction solution, 0.5M (458) and 1.0M Na2SO4 (474) were experimented with as the reaction solvent. When 0.5 M Na2SO4 was used there was an increase in the amount of PVP that was fixed, 154 mg/mL for 458 compared to 120 mg/mL for 410. When the sodium sulfate concentration was increased to 1.0 M very less PVP was fixed, only 79 mg/mL for 474. However, for both syntheses there was a pronounced precipitation of PVP in the reaction solution that probably prevented effective coupling.

[0085] For the following experiment 493, in order to avoid precipitation a lower amount of PVP was tried, only 10% while keeping 1 M Na2SO4 as reaction solvent. This proved to be a successful medium from then on. 192 mg/mL of PVP was fixed in this experiment. A slight phase separation with small polymer particles can be seen, but when water is added after the reaction is complete the polymer droplets are rapidly dissolved and the reaction solution becomes clear. Particles can be quickly washed away in a filter. Polymer droplets should be kept under observation during scale to make sure they are easily dissolved.

[0086] A higher ratio of VP versus particles (ie, a lower concentration of reaction slurry) was also experienced on the 518 with in all respects the same graft formula as on the 493. For 518 the amount of particles was 50 mL compared to 100 mL that were used for all other syntheses. It was considered that this would increase the amount of PVP further, but this was not the case, it had less PVP fixed with this formula, 116 mg/mL for 518 compared to 194 mg/mL for 493.

[0087] 538 is a reproduction of 493 to ensure the formula is robust. An almost identical result was obtained at 194 mg/mL in PVP set for 538 compared to 192 mg/mL for 493.

[0088] A final experiment, 796, was done in order to see the effect of allyl content on the optimized graft formula. For 796 an allyl level of 82 μmol/mL was used compared to 166 μmol/mL which was used for 493. This resulted in a lower amount of fixed PVP 165 mg/mL compared to 192 mg/mL for 493. Alila level, however is not drastic. • Boiling can be avoided for experiments where 19.6% PV or less is used. • Temperature has an effect on the large amount of PVP that is set and on the tendency to boil. Temperatures above 35°C improve the efficiency of the coupling reaction. • The amount of VP+primer is an important parameter of the synthesis; it significantly affects the amount of PVP that is fixed, the tendency to boil and the viscosity of the reaction solution. • The allyl content affects the amount of PVP that is fixed; higher allyl contents give higher amounts of fixed PVP. The effect, however, is not dramatic. When the amount of allyl groups is reduced from 166 to 82 μmol/mL the amount of fixed PVP is reduced only from 192 to 165 mg/mL. • The use of sodium sulfate in the reaction solution has a large positive effect on the amount of PVP that is fixed, as long as precipitation of PVP is avoided. • Boiling, a viscous reaction solution or precipitation of PVP were avoided for prototype 493 where 10% VP in combination with 1.0 M Na2SO4 was used. These conditions resulted in 192 mg/mL of PVP fixed which is similar to the amount of PVP that was fixed for the reference prototype 9018, 199 mg/mL. This prototype performed well, and the reaction conditions used for this prototype can be used as a starting point for further optimization.

Tabela 3. Protótipos avaliados com o processo de estabilização de cerveja[0089] The prototypes, shown in Table 3, except for 9018, were evaluated by flowing 750 ml of unstabilized lager beer through a column filled with 1 ml of the resin prototype at a flow rate of 3.3 ml/min (18 seconds of residence time) and at a temperature of <5°C. Beer flow through was analyzed with "Alcohol Cold Cloudiness", "Tanoids" and "Total Polyphenols" and compared to unstabilized beer and unstabilized beer that had been processed through CSS adsorber (Q Sepharose BB). For the first three prototypes tested the Tanoid content was zero and the Tanoid content test was left to the rest of the prototypes as more beer would need to be processed through the 1 mL column to assess the thanoid breakdown. It should be noted that a high level of tanoids was determined in the beer after processing in the CSS adsorber, giving an indication that the prototypes are working extremely well as the "tanoids" assay is considered to be the best correlated assay for colloidal stability.

Table 3. Prototypes evaluated with the beer stabilization process

Beer stabilization methods Column filling:

[0090] Each prototype and reference (CSS adsorber lot 10039019) were packed into a Tricorn 5/100 column (GE Healthcare) at 2 mL/min. The bed height was adjusted to 5.1 ± 0.1 cm at 2 mL/min to obtain a bed volume of 1.0 mL of resin. The top adapter was set 1 mm below the 5.1 mark and the column was equilibrated at 2 mL/min with 5 column volumes of water prior to application of beer.

beer application

[0091] Figure 4 shows the adjustment of the miniaturized beer stabilization application. An 18 L Cornelius bottle of unstabilized beer was obtained from the Slottkallans brewery (Uppsala, Sweden) and placed inside an incubator at 0°C for three days. New beer that has been sterile filtered must be reconstituted over a few days to obtain stable results. The beer bottle tube was divided into 2 pumps and 4 pump heads (P-900 pumps, GE Healthcare) to be able to run 4 prototypes simultaneously. Since it is impossible to remove bubbles from beer tubes the actual flow rate must be calibrated. All pump heads were set to 4.0 mL/min and 10 mL volumetric flasks were filled and the time to fill the volumetric flasks to the mark was observed and the actual flow rate with beer was calculated to be 3.3 mL/min for all pump heads. The 4 columns with prototypes were placed inside the incubator and the tubes from the pump heads were connected to the columns. 1000 mL collection bottles were attached after each column in the incubator. The first column in the test series was always the reference column containing CSS adsorber lot 10039019. These beads had diameters in the range of 100-300 micrometers, with a volume-weighted average diameter of 200 micrometers and a quaternary ammonium group content of 0.18 -0.25 mmol/ml. 750 mL of beer was pumped through each column at a flow rate of 3.3 mL/min which corresponds to a residence time of 18 s, ie ~3 times faster than the common CSS process. The process time was 3.75 h. Beer that had flowed through the column was collected in the 1000 mL bottles.

Alcohol turbidity analysis methods

[0092] Beer samples from the 1000 mL collection bottles were analyzed for 20 hours after the beer stabilization process. First, unstabilized beer from the Cornelius bottle was analyzed followed by the CSS adsorber reference sample and prototypes. ~20 mL of beer was transferred into a 50 mL Falcon tube and the tube was shaken rapidly to remove carbon dioxide. After the beer had settled, 4 x 1.0 mL of beer was carefully pipetted into the cuvette. 120 μL of ethanol was pipetted into a clean cuvette and the cuvette was carefully turned up and down 5 times before analysis. 0.6 mL of ethylene glycol was added to the cuvette chamber in the Tanometer to increase the thermal contact between the cuvette and the cooler. The cuvette was placed in the cuvette chamber and the alcohol cold turbidity analysis was started. The sample in the cuvette was frozen to -5°C and turbidity was measured after the sample had been incubated for 40 minutes. It was shown that after ~10 minutes, the turbidity remained the same as during the 40 minutes, so for some samples the turbidity was monitored after 20 minutes to speed up the analysis.

[0093] The cold turbidity of alcohol was calculated by subtracting the final turbidity from the initial turbidity in EBC units. See figure 5.

Tanoid Methods

[0094] A perfusion syringe, made of glass, was filled with 0.400 g/L of the PVP solution and placed on the support provided in the Tanometer (Pfeuffer GmbH, Germany). Beer samples from the 1000 mL collection bottles were analyzed within 20 hours after the beer stabilization process. First, unstabilized beer from the Cornelius bottle was analyzed followed by the reference sample from the CSS adsorber and the prototypes. ~20 mL of beer was transferred into a 50 mL Falcon tube and the tube was shaken rapidly to remove carbon dioxide. After the beer had settled, 4 x 1.0 mL of the beer was carefully pipetted into the cuvette. A stirring rod was placed in the cuvette bottle. The sample was titrated with PVP solution from the perfusion syringe at 5 mL/h at 25°C until 100 mg/L PVP or until the tanoid peak reached its maximum and the tanoid content was automatically calculated by the Tannolab software.

Total polyphenols methods

[0095] Beer samples from the 1000 mL collection bottles were analyzed within 4 hours after the beer stabilization process. First, unstabilized beer from the Cornelius bottle was analyzed followed by the CSS adsorber reference sample and prototypes. ~50 mL of beer was filtered through a Whatman™ filter paper into a 200 mL E-flask. 2 x 10 mL of beer were pipetted into a 25 mL volumetric flask. 8 ml of the CMC/EDTA solution was added into volumetric flasks. 500 μl of ferric reagent was added to one vial only and 500 µl of ammonia solution was pipetted into both volumetric vials. MilliQ™ water was added to the volumetric flask brand. The vials were mixed quickly. The vial without ferric reagent was the blank sample. Absorbance at 600 nm was measured after >10 min (within 60 min) for sample and blank, using a 10 cm cuvette. The total polyphenol in the vial was calculated by the formula: TP = (As-Ab) x 820 Where TP = Total Polyphenols (mg/L) As = Absorbance for the sample, AU AB = Absorbance for blank, AU

[0096] The total polyphenol reduction was calculated by dividing the total polyphenols from the processed beer sample across the prototypes with the total polyphenols from the unstabilized beer.

[0097] Since the chemical and physical properties of unstabilized beer change rapidly, it was only possible to operate and analyze prototypes 3-6 and reference CSS adsorber with unstabilized beer within the required time period of 24 hours. For this reason, the 12 prototypes were worked with the application of beer in 4 cycles.

[0098] First the prototypes were analyzed with the Tanoid content test. Since the first three prototypes bound all the tanoids in the beer, it was impossible to assess their ability to bind tanoids, so the following prototypes were evaluated only with the analysis of cold turbidity of alcohol. For PVP-grafted beads, a larger amount of beer per half mL may be needed to investigate tanoid capacity. However, the CSS adsorber beads only reduced the tanoid content from 69.8 to 32.8 mg/L, indicating that the PVP grafted prototypes are significantly more capable of binding tanoids than the CSS adsorber.

[0099] Figure 6 shows the results of the tanoid content. Lot "Uppsala 1" was used.

[00100] Table 4 shows the results of the alcohol cold turbidity analysis for the prototypes. Cold haze was normalized and the relative percentage of cold haze compared to unstabilized beer was calculated for the prototypes. Lot "Uppsala 1" was used. Table 4. Results of alcohol cold turbidity.

[00101] To estimate the variance within the assay to judge whether the results differ significantly or not, the mean and relative standard deviation of the CSS adsorber reference that was run four times was evaluated. N = 4 Mean = 42.1% SD = 2.9% RSD = 6.9%

- Todos os protótipos mostraram menos turvação a frio do que o adsorvedor CSS usando aproximadamente os mesmos ajustes de processo de cerveja que os do processo do adsorvedor CSS usa hoje, 750 mL cerveja/meio mL. - Dois protótipos mostraram divergências fortemente significantes em relação a outros protótipos. Cerveja processada através de 392 não mostrou turvação a frio de modo algum e 474 mostrou uma quantidade superior fortemente significante de turvação a frio em relação aos outros protótipos. - Cerveja processada através de todos os outros protótipos mostraram uma turvação a frio de 1,4-4,4 unidades EBC e elas foram difíceis de diferenciar. - Todos os protótipos mostraram redução superior do polifenol total em relação ao adsorvedor CSS. O protótipo 474 mostrou menos redução no polifenol total em relação aos outros protótipos.[00102] Table 5 and Figure 8 show the results of total polyphenol reduction. The total polyphenol reduction was measured for the 750 mL beer fraction and also for the 50 mL fraction, collected after the 750 mL to investigate if the prototype was still absorbing the total polyphenols. In figure 6, for each prototype the first bar is the 750 mL fraction and the second bar is the 50 mL fraction. Beer batch "Uppsala 2" was used. Table 5. Results of polyphenol reduction.

- All prototypes showed less cold haze than the CSS adsorber using approximately the same beer process settings as the CSS adsorber process uses today, 750 mL beer/half mL. - Two prototypes showed strongly significant differences in relation to other prototypes. Beer processed through 392 showed no cold haze at all and 474 showed a significantly higher amount of cold haze compared to the other prototypes. - Beer processed through all other prototypes showed a cold cloudiness of 1.4-4.4 EBC units and they were difficult to differentiate. - All prototypes showed superior reduction of total polyphenol in relation to CSS adsorber. Prototype 474 showed less reduction in total polyphenol than the other prototypes.

[00103] It should be noted that the alcohol turbidity test also includes simple monomeric flavanols that are not capable of crosslinking peptides above 0°C and with less alcohol content. These flavanols are also unable to polymerize as the oxidation products of these polyphenols are stable. Beer processed by 311, 377 and 431 showed zero tanoid content and differed by 2.8-4.4 EBC units in the cold turbidity. It is for this reason mainly the selectivity to bind simple monomeric flavanols that differs between the prototypes.

[00104] From the perspective of the prototype design, to judge how the results of the beer application results relate to the parameters used in the synthesis there is in general no correlation except for prototype 392 (no cold turbidity was observed) and 474 (relatively high cold turbidity content). 392 is the only prototype that was synthesized at 35°C, 10-20°C lower than the other prototypes. It can be speculated that ligand synthesis at higher temperature affects the formability of the PVP polymer and at lower temperature the polymer is more flexible and able to bind even low molecular weight polyphenols with few exposed hydroxyl groups. 474 had a much lower binder density than the other prototypes and the breakdown of more complex polyphenols occurred during the process. Figure 8 shows the correlation between total polyphenols and cold turbidity. From this figure, it is easy to distinguish between prototypes that have >100 mg PVP/half mL and high polyphenol reduction compared to a prototype with low PVP/mL (474) medium and CSS adsorber. Examples 2. Comparative Examples

[00105] Diethylene glycol vinyl ether (DEGVE) grafted agarose beads were prepared according to the methods described in US20100028505 and US Pat. 8,137,559 by alkylating Sepharose™ 6 FastFlow crosslinked agarose beads (GE Healthcare BioSciences AB) with allyl glycidyl ether and reacting 10 g of wet allylated beads with a solution of 1.6 g of 2,2'-azobis (2-methylbutyronitrile) in 40 g of diethylene glycol vinyl ether at 70°C for 18 h under an inert atmosphere. The beads were then washed with large amounts of water and ethanol. Grafted DEGVE content was measured from increasing dry content and found to be 0.76 mmol of DEGVE monomer residues per mL of beads.

[00106] 5 ml aliquots of the DEGVE grafted prototype and a sample of CSS adsorber beads were filled into XK 16 columns (GE Healthcare Bio-Sciences AB) and 1000 ml of filtered unstabilized beer was pumped through each column at a flow rate of 13 ml/min. Cold turbidity was measured before and after passage through the columns and the reduction in turbidity was verified to be properly the same as for the DEGVE prototype and the CSS adsorber. Consequently, as in the previous example, the PVP grafted prototypes showed significantly better haze reduction than the CSS adsorber beads, they are also better than the DEGVE prototype.

[00107] This written description uses examples to describe the invention, including the best mode, and also to enable anyone skilled in the art to practice the invention, including making and using any devices or systems and carrying out any incorporated methods. The patentable scope of the invention is defined by the claims, and may include examples that occur to those skilled in the art. Such other examples are intended to be within the scope of the claims if they have structural elements that do not differ from the literal language of the claims, or if they include equivalent structural elements with insubstantial differences from the literal language of the claims. All patents and patent applications mentioned in this text are incorporated herein by reference in their entirety as if they were individually incorporated.

Claims (15)

1. Separation matrix comprising a porous solid support and a plurality of polyvinylpyrrolidone polymer chains covalently attached to said solid support, characterized in that said polyvinylpyrrolidone polymer chains are copolymer chains or vinylpyrrolidone homopolymer chains comprising at least 70% by mole of vinylpyrrolidone monomer residues and less than 2% by moles of negatively charged monomer residues, wherein said porous solid support comprises a polysaccharide selected from the group consisting of agarose, agar, cellulose and dextran, and in that said separation matrix comprises 100-200 polyvinylpyrrolidone polymer per ml of matrix. 2. Separation matrix according to claim 1, characterized in that said polyvinylpyrrolidone polymer chains comprise: a) less than 1% by mol of negatively charged monomer residues; and/or b) up to 30% in moles of positively charged monomer residues. 3. Separation matrix according to claim 1 or 2, characterized in that said solid porous support: a) comprises particles, such as particles having an average diameter of 10-500 micrometers; and/or b) has a porosity of 80-98%, such as 90-98%. 4. Separation matrix according to any one of claims 1 to 3, characterized in that a plurality of said polyvinylpyrrolidone polymer chains are each covalently bonded to said solid support by a single binding moiety, such as in which said single binding moiety comprises an ether-linked C3 chain. 5. Separation matrix according to any one of claims 1 to 4, characterized in that it comprises: 0.5-4.0, such as 0.7-3.0 micromoles of vinylpyrrolidone monomer residues per ml of headquarters; and/or 0.50-0.80 g of polyvinylpyrrolidone polymer per g of dry matrix. 6. Separation matrix according to any one of claims 1 to 5, characterized in that it comprises 120-180 mg of polyvinylpyrrolidone polymer per ml of matrix; and/or less than 5 micrograms/g of leachable carbon. 7. Method for stabilizing a fermented beverage, characterized in that it comprises the steps of: a) providing a column filled with a separation matrix as defined in claim 1, and b) passing said beverage through said column and recovering a flow passing through the spine like a stabilized drink. 8. Method according to claim 7, characterized in that the separation matrix is defined as any one of claims 2 to 6 9. Method according to claim 7 or 8, characterized in that it comprises steps c) of column regeneration with a regeneration solution and; d) repeating steps a) - c) at least twice, such as at least 10 or at least 50 times, wherein said regeneration solution optionally comprises NaOH, such as at least 0.1 M NaOH or 0, 1-2M NaOH. 10. Method according to any one of claims 79, characterized in that in step b) the residence time of the beverage in the column is 2 minutes or less, such as 1 minute or less. 11. Method according to any one of claims 7 to 10, characterized in that the fermented beverage is beer, and wherein the beverage stabilized in step b) has a cold haze of less than 25% of the cold haze for the beverage before passing through said column. 12. Method for manufacturing the separation matrix as defined in any one of claims 1 to 6, characterized in that said method comprises the steps of: a) providing a porous solid support comprising at least 5 micromol/ml of reactive moieties to radical, wherein said solid porous support comprises a polysaccharide selected from the group consisting of agarose, agar, cellulose and dextran; b) contacting said support with a monomer composition wherein at least 70% of the monomers in said monomer composition are N-vinyl pyrrolidone; c) initiating free radical polymerization to form a matrix having polyvinylpyrrolidone polymer chains covalently attached to the support, and; d) washing said matrix. 13. Method according to claim 12, characterized in that said radical-reactive fractions are C=C double bonds. Method according to claim 12 or 13, characterized in that step a) comprises i) providing a divinylbenzene copolymer support or ii) reacting a hydroxyfunctional support with an allyl halide or allylglycidyl ether. Method according to any one of claims 12 to 14, characterized in that step c) is carried out with the support suspended in an aqueous solution comprising N-vinyl pyrrolidone and 0.05-3 mol/L of a salt , such as 0.1 - 2 M sodium sulfate.

Images